为解决沉管隧道管节长度传统优化算法在多目标优化中的局限性问题，提出一种基于非支配排序的改进鲸鱼优化算法（NDSWOA）。首先，分析沉管隧道管节长度对施工工期、造价及结构安全性的影响因素，并建立多目标优化模型; 然后，设计并实现NDSWOA算法，将非支配排序机制与鲸鱼优化算法相结合，以增强算法的全局搜索能力和解集多样性; 最后，选取深圳沿江沉管隧道项目为工程案例，采用NDSWOA优化管节长度，并与NSGA-II和MOEA/D等传统优化算法进行对比分析。试验结果表明： 1）NDSWOA在处理沉管隧道管节长度优化问题时，展现出更快的收敛速度和更优的解质量，能够生成均匀分布的帕累托前沿解。2）相比于NSGA-II和MOEA/D，NDSWOA在优化工期、造价和结构安全性方面表现更优，由此优化得到推荐管节长度为80 m，可在各优化目标之间取得平衡。

To address the limitations of traditional algorithms in the multiobjective optimization of immersed tunnel segment lengths, an improved algorithm based on nondominated sorting whale optimization algorithm (NDSWOA) is proposed. First, the factors influencing construction duration, cost, and structural safety associated with segment length in immersed tunnels are analyzed, and a multiobjective optimization model is established. Second, the NDSWOA is designed and implemented by integrating nondominated sorting mechanism into the whale optimization algorithm to enhance global search capability and solution diversity. Finally, the NDSWOA is applied to optimize segment length in a case study of the Yanjiang immersed tunnel in Shenzhen, China. The results are compared with those obtained using traditional optimization algorithms such as NSGA-Ⅱ and MOEA/D. The experimental results reveal the following: (1) The NDSWOA demonstrates faster convergence speed and superior solution quality in solving the immersed tunnel segment length optimization problem, effectively generating a well-distributed Pareto front. (2) Compared with NSGA-Ⅱ and MOEA/D, the NDSWOA achieves better performance in optimizing construction duration, cost, and structural safety, suggesting an 80 m segment length as an optimal trade-off among the optimization objectives.